Turk's theory of norm resistance explains how authority-subject relations can be structured in manners that have different probabilities of overt conflict (norm resistance). Building on previous research by Lanza-Kaduce and Greenleaf, this study uses data collected as part of an observational study of the police in Indianapolis, Indiana, and St. Petersburg, Florida, to examine Turk's theory as it relates to conflict in police-suspect encounters. It examines three hypotheses derived from the theory of norm resistance, using multivariate statistical techniques to control for several factors either posited or empirically shown in previous research to influence overt conflict. Two of the three hypotheses are supported. Consistent with prior research, organization and sophistication of police and suspects are significant predictors of overt conflict. However, the hypothesis that conflict will be less likely when officers' positional authority is reinforced by race, age, sex, and wealth deference norms is not supported. (Contains 3 tables and 19 notes.)